将行中的数据添加到R中的列

Question

I have the following table:

     wk Brand Retail_price
1     1     a          1.2
2     2     a          1.3
3     1     c          1.4
4     2     c          1.5
5     1     d          1.6
6     2     d          1.7

I am trying to get the retail price of other brands during the same week as columns. I want to get to:

     wk Brand Retail_price Retail_price_a Retail_price_c Retail_price_d
1     1     a          1.2             NA            1.4            1.6
2     2     a          1.3             NA            1.5            1.7
3     1     c          1.4            1.2             NA            1.6
4     2     c          1.5            1.3             NA            1.7
5     1     d          1.6            1.2            1.4             NA  
6     2     d          1.7            1.3            1.5             NA 

I have tried looping through the data to manually add columns for each brand. Turned out to be very inefficient.

I am new to R. I am thinking of something equivalent to pd.pivot in python, create a new df and then join the two.

How can I do this in R? Is there a better way of doing this?

Answer 1

I think you need a combination of reshaping (long-to-wide) and merge. Here's an example using dplyr and tidyr :

# data
x <- data.frame(
  wk = c(1L, 2L, 1L, 2L, 1L, 2L),
  Brand = c("a", "a", "c", "c", "d", "d"),
  Price = c(1.2, 1.3, 1.4, 1.5, 1.6, 1.7),
  stringsAsFactors = FALSE)

library(dplyr)
library(tidyr)
x2 <- spread(x, Brand, Price, sep = "_") %>%
  left_join(x, by = "wk")
x2
#   wk Brand_a Brand_c Brand_d Brand Price
# 1  1     1.2     1.4     1.6     a   1.2
# 2  1     1.2     1.4     1.6     c   1.4
# 3  1     1.2     1.4     1.6     d   1.6
# 4  2     1.3     1.5     1.7     a   1.3
# 5  2     1.3     1.5     1.7     c   1.5
# 6  2     1.3     1.5     1.7     d   1.7

You can then work on removing same-brand cells if needed.

In data.table :

library(data.table)
xDT <- setDT(copy(x))

merge(xDT, dcast(xDT, wk ~ Brand), by = c("wk"))
# or #
xDT[dcast(xDT, wk ~ Brand), on = c("wk")]